Marvell1
Legendary
Offline
Activity: 2576
Merit: 1136
Leading Crypto Sports Betting & Casino Platform
|
|
June 16, 2014, 07:42:39 PM |
|
Love the design of these miners, so dense with asic chips, but the board arrangement of the heatsinks and fans are just perfect. These guys just know what they're doing. Wish I had $5k to throw at that August SP30 Group buy.... I wish i have 5k for the sept ones
|
..Stake.com.. | | | ▄████████████████████████████████████▄ ██ ▄▄▄▄▄▄▄▄▄▄ ▄▄▄▄▄▄▄▄▄▄ ██ ▄████▄ ██ ▀▀▀▀▀▀▀▀▀▀ ██████████ ▀▀▀▀▀▀▀▀▀▀ ██ ██████ ██ ██████████ ██ ██ ██████████ ██ ▀██▀ ██ ██ ██ ██████ ██ ██ ██ ██ ██ ██ ██████ ██ █████ ███ ██████ ██ ████▄ ██ ██ █████ ███ ████ ████ █████ ███ ████████ ██ ████ ████ ██████████ ████ ████ ████▀ ██ ██████████ ▄▄▄▄▄▄▄▄▄▄ ██████████ ██ ██ ▀▀▀▀▀▀▀▀▀▀ ██ ▀█████████▀ ▄████████████▄ ▀█████████▀ ▄▄▄▄▄▄▄▄▄▄▄▄███ ██ ██ ███▄▄▄▄▄▄▄▄▄▄▄▄ ██████████████████████████████████████████ | | | | | | ▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▄ █ ▄▀▄ █▀▀█▀▄▄ █ █▀█ █ ▐ ▐▌ █ ▄██▄ █ ▌ █ █ ▄██████▄ █ ▌ ▐▌ █ ██████████ █ ▐ █ █ ▐██████████▌ █ ▐ ▐▌ █ ▀▀██████▀▀ █ ▌ █ █ ▄▄▄██▄▄▄ █ ▌▐▌ █ █▐ █ █ █▐▐▌ █ █▐█ ▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀█ | | | | | | ▄▄█████████▄▄ ▄██▀▀▀▀█████▀▀▀▀██▄ ▄█▀ ▐█▌ ▀█▄ ██ ▐█▌ ██ ████▄ ▄█████▄ ▄████ ████████▄███████████▄████████ ███▀ █████████████ ▀███ ██ ███████████ ██ ▀█▄ █████████ ▄█▀ ▀█▄ ▄██▀▀▀▀▀▀▀██▄ ▄▄▄█▀ ▀███████ ███████▀ ▀█████▄ ▄█████▀ ▀▀▀███▄▄▄███▀▀▀ | | | ..PLAY NOW.. |
|
|
|
bobsmoke
|
|
June 16, 2014, 08:17:39 PM |
|
Allied Control (DataTank) experimenting with SP10 immersion cooling:
WOW !!!
|
|
|
|
Tsengar
Newbie
Offline
Activity: 9
Merit: 0
|
|
June 17, 2014, 04:38:47 AM |
|
Allied Control (DataTank) experimenting with SP10 immersion cooling: clip1pic1pic2pic3Good luck in the upcoming Havelock IPO Disclaimer: We don't endorse the IPO. Each investor should do his/her ROI calculation and decide whether to invest or not. That is interesting video right there... Those chips don't seem to get the Novec vapourizing much. Speaks well to the SP10 as an air cooled unit. SP30 board be keen to see that. Quick introduction, since this will be my fist post here. I am a work colleague of antirack, and I am the one that have been doing the setups for this "lab" tests. I have been running the SP10 for quite a while now, it works great, just not very optimal circumstances for the boards as they had to be placed horizontal and heat dissipation side down. Hence also a bit tricky to get a good recording of the action, if you so will.
|
|
|
|
Guy Corem (OP)
Donator
Legendary
Offline
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
|
|
June 17, 2014, 04:53:47 AM Last edit: June 17, 2014, 05:07:03 AM by Spondoolies-Tech |
|
Allied Control (DataTank) experimenting with SP10 immersion cooling: clip1pic1pic2pic3Good luck in the upcoming Havelock IPO Disclaimer: We don't endorse the IPO. Each investor should do his/her ROI calculation and decide whether to invest or not. That is interesting video right there... Those chips don't seem to get the Novec vapourizing much. Speaks well to the SP10 as an air cooled unit. SP30 board be keen to see that. Quick introduction, since this will be my fist post here. I am a work colleague of antirack, and I am the one that have been doing the setups for this "lab" tests. I have been running the SP10 for quite a while now, it works great, just not very optimal circumstances for the boards as they had to be placed horizontal and heat dissipation side down. Hence also a bit tricky to get a good recording of the action, if you so will. Short explanation: Unlike AM Gen3, we designed the our 1st gen ASIC (Hammer) to dissipate the heat downward, which is far better for QFN package, in our opinion.
|
|
|
|
dogie
Legendary
Offline
Activity: 1666
Merit: 1185
dogiecoin.com
|
|
June 17, 2014, 05:24:05 AM |
|
Allied Control (DataTank) experimenting with SP10 immersion cooling: clip1pic1pic2pic3Good luck in the upcoming Havelock IPO Disclaimer: We don't endorse the IPO. Each investor should do his/her ROI calculation and decide whether to invest or not. That is interesting video right there... Those chips don't seem to get the Novec vapourizing much. Speaks well to the SP10 as an air cooled unit. SP30 board be keen to see that. Quick introduction, since this will be my fist post here. I am a work colleague of antirack, and I am the one that have been doing the setups for this "lab" tests. I have been running the SP10 for quite a while now, it works great, just not very optimal circumstances for the boards as they had to be placed horizontal and heat dissipation side down. Hence also a bit tricky to get a good recording of the action, if you so will. Short explanation: Unlike AM Gen3, we designed the our 1st gen ASIC (Hammer) to dissipate the heat downward, which is far better for QFN package, in our opinion. Isn't it a mistake to design a QFN package to do anything but (mostly) dissipate downwards?
|
|
|
|
Guy Corem (OP)
Donator
Legendary
Offline
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
|
|
June 17, 2014, 05:30:49 AM |
|
Allied Control (DataTank) experimenting with SP10 immersion cooling: clip1pic1pic2pic3Good luck in the upcoming Havelock IPO Disclaimer: We don't endorse the IPO. Each investor should do his/her ROI calculation and decide whether to invest or not. That is interesting video right there... Those chips don't seem to get the Novec vapourizing much. Speaks well to the SP10 as an air cooled unit. SP30 board be keen to see that. Quick introduction, since this will be my fist post here. I am a work colleague of antirack, and I am the one that have been doing the setups for this "lab" tests. I have been running the SP10 for quite a while now, it works great, just not very optimal circumstances for the boards as they had to be placed horizontal and heat dissipation side down. Hence also a bit tricky to get a good recording of the action, if you so will. Short explanation: Unlike AM Gen3, we designed the our 1st gen ASIC (Hammer) to dissipate the heat downward, which is far better for QFN package, in our opinion. Isn't it a mistake to design a QFN package to do anything but (mostly) dissipate downwards? Gross mistake, IMHO
|
|
|
|
Tsengar
Newbie
Offline
Activity: 9
Merit: 0
|
|
June 17, 2014, 05:33:58 AM |
|
Allied Control (DataTank) experimenting with SP10 immersion cooling: clip1pic1pic2pic3Good luck in the upcoming Havelock IPO Disclaimer: We don't endorse the IPO. Each investor should do his/her ROI calculation and decide whether to invest or not. That is interesting video right there... Those chips don't seem to get the Novec vapourizing much. Speaks well to the SP10 as an air cooled unit. SP30 board be keen to see that. Quick introduction, since this will be my fist post here. I am a work colleague of antirack, and I am the one that have been doing the setups for this "lab" tests. I have been running the SP10 for quite a while now, it works great, just not very optimal circumstances for the boards as they had to be placed horizontal and heat dissipation side down. Hence also a bit tricky to get a good recording of the action, if you so will. Short explanation: Unlike AM Gen3, we designed the our 1st gen ASIC (Hammer) to dissipate the heat downward, which is far better for QFN package, in our opinion. I am pretty sure same goes for the BE200 ASIC, but I leave that up to you engineer guys. My comment was of course not meant be negative in any way, just explaining the situation of this tests. I also intentionally kept the ASCI side up to make it look more interesting then a big copper plate. Hopefully soon I can show some boiling action of both sides in a better way.
|
|
|
|
Guy Corem (OP)
Donator
Legendary
Offline
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
|
|
June 17, 2014, 05:35:26 AM |
|
Disclaimer: We don't really know what AM designed. We're just looking at the IC video published by AC
|
|
|
|
jtoomim
|
|
June 17, 2014, 06:46:34 AM |
|
Hey Spondoolies, I read this suggestion on Reddit a little bit ago, and thought it sounded promising as a method to reduce the probability of certain types of 51%-type attacks (*cough* ghash.io). The basic idea is to include a SPV client in the miner, and by default to refuse to mine if the pool gives a block header that isn't built on the longest published blockchain. I'd be curious to hear what you guys (especially Adam Beck and Zvi Shteingart) think about this idea in terms of the efficacy of this approach and the ease of implementation. Obviously, this wouldn't work well for people wanting to mine altcoins like peercoin, but a simple checkbox could fix that. Most miners are lazy and greedy, and just want to plug and play something that makes them money, which means they probably take care of the security of the blockchain unless their manufacturers do it for them. http://www.reddit.com/r/Bitcoin/comments/2873ox/making_pooled_mining_immune_to_51_attacks_selfish/
|
Hosting bitcoin miners for $65 to $80/kW/month on clean, cheap hydro power. http://Toom.im
|
|
|
maardein
|
|
June 17, 2014, 06:58:16 AM |
|
Allied Control (DataTank) experimenting with SP10 immersion cooling: clip1pic1pic2pic3Good luck in the upcoming Havelock IPO Disclaimer: We don't endorse the IPO. Each investor should do his/her ROI calculation and decide whether to invest or not. Immersion cooling always looks so cool Keep up the experimenting!
|
BTC: 1788UegKXGXXicfPcbZ1bmSUJ99ZWRCF7p LTC: LZ2rCcoxK4X8wRRynqdxoimd4d3TDNk7Lk PMP: PApSSdorQds5tQysymwDXPAN3viJLFTUs8
|
|
|
Bicknellski
|
|
June 17, 2014, 06:58:48 AM |
|
Allied Control (DataTank) experimenting with SP10 immersion cooling: clip1pic1pic2pic3Good luck in the upcoming Havelock IPO Disclaimer: We don't endorse the IPO. Each investor should do his/her ROI calculation and decide whether to invest or not. That is interesting video right there... Those chips don't seem to get the Novec vapourizing much. Speaks well to the SP10 as an air cooled unit. SP30 board be keen to see that. Quick introduction, since this will be my fist post here. I am a work colleague of antirack, and I am the one that have been doing the setups for this "lab" tests. I have been running the SP10 for quite a while now, it works great, just not very optimal circumstances for the boards as they had to be placed horizontal and heat dissipation side down. Hence also a bit tricky to get a good recording of the action, if you so will. Short explanation: Unlike AM Gen3, we designed the our 1st gen ASIC (Hammer) to dissipate the heat downward, which is far better for QFN package, in our opinion. I am pretty sure same goes for the BE200 ASIC, but I leave that up to you engineer guys. My comment was of course not meant be negative in any way, just explaining the situation of this tests. I also intentionally kept the ASCI side up to make it look more interesting then a big copper plate. Hopefully soon I can show some boiling action of both sides in a better way. You could see a lot of action (vapor bubbles) escaping through the mounting holes and out from under the edges of the board as the vapor builds up under the board it makes for an interesting show but obviously this is not the normal configuration. If it were a vertical setup you would have that rolling "boil" look I suppose. It just bet the SP10 board so well designed that in comparison to a smaller pcb of a higher density with 'older' hotter chips is comparing apples to dragonfruit with the SP10. They don't really need to be in Novec with those Hammer chips, but if you had a much smaller blade maybe that would be required. It be great to get some numbers on the SP10 especially can you overclock the snot out of the Hammer chips in the Novec even laying flat? Or do we all have to wait for Inside Bitcoin Hong Kong to learn more about the potential? I can't wait to get to Allied Control and check all these things out. Curiosity!
|
|
|
|
Collider
|
|
June 17, 2014, 07:35:53 AM |
|
I think it is a good sign that AC is experimenting with these designs.
However, the first really profitable implementation should be with the sp30s, as they are much more power efficient.
|
|
|
|
Guy Corem (OP)
Donator
Legendary
Offline
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
|
|
June 17, 2014, 08:06:26 AM |
|
It be great to get some numbers on the SP10 especially can you overclock the snot out of the Hammer chips in the Novec even laying flat? Or do we all have to wait for Inside Bitcoin Hong Kong to learn more about the potential? I can't wait to get to Allied Control and check all these things out.
Curiosity!
AC will publish the ASIC page soon. Take into account that it's just initial testing and not tuned for performance yet.
|
|
|
|
Guy Corem (OP)
Donator
Legendary
Offline
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
|
|
June 17, 2014, 08:06:56 AM |
|
I think it is a good sign that AC is experimenting with these designs.
However, the first really profitable implementation should be with the sp30s, as they are much more power efficient.
Indeed.
|
|
|
|
Tsengar
Newbie
Offline
Activity: 9
Merit: 0
|
|
June 17, 2014, 09:14:08 AM |
|
Some ASIC stats after playing around with it for a while, can most likely push it a bit further, enjoy. Once again, this is in no way optimal conditions for the boards, with the heat dissipation surface down. http://i59.tinypic.com/2qan7kw.png
|
|
|
|
Collider
|
|
June 17, 2014, 09:16:36 AM Last edit: June 17, 2014, 09:59:05 AM by Collider |
|
Some ASIC stats after playing around with it for a while, can most likely push it a bit further, enjoy. Once again, this is in no way optimal conditions for the boards, with the heat dissipation surface down.
Is that in an immersion cooled environment? Did you switch the PSU? I am getting around the same speed at 1240W... and with 8 inactive chips.... Temps are looking good though. Is yours a May batch unit? Also: As you can see the chips are clocking at different speeds although they all show the same temperature. This is most likely due to the firmware anticipating higher temperatures at the back of the unit, as you would typically see with air cooling. The unit should be running a little more efficient (and possibly faster) if all clock speeds rise equally.
|
|
|
|
zvisha
|
|
June 17, 2014, 10:18:27 AM |
|
Some ASIC stats after playing around with it for a while, can most likely push it a bit further, enjoy. Once again, this is in no way optimal conditions for the boards, with the heat dissipation surface down.
Can not compare slow corner May/July to early fast corner March/April. If it was fast corner it would be over 1.6T for sure. Is 1300W the limit you discovered by trial (after removing fans)?
|
|
|
|
Tigggger
Legendary
Offline
Activity: 1098
Merit: 1000
|
|
June 17, 2014, 10:27:05 AM |
|
Spondoolies
Can you provide some more information on the hosting arrangements for the SP30 group buy, still sat on the email asking if I want it shipped or hosted, was going to email you but figured some other people might be in the same position so a public answer would be better.
It says 6 month minimum, how is this going to be charged. All 6 months in advance, 1 month in advance, 1 month in arrears.
Will we have remote access to the miners, my current SP10's are hosted with you and the biggest drawback is not having access (though support is very quick)
TIA
|
|
|
|
zvisha
|
|
June 17, 2014, 12:38:45 PM |
|
Hey Spondoolies, I read this suggestion on Reddit a little bit ago, and thought it sounded promising as a method to reduce the probability of certain types of 51%-type attacks (*cough* ghash.io). The basic idea is to include a SPV client in the miner, and by default to refuse to mine if the pool gives a block header that isn't built on the longest published blockchain. I'd be curious to hear what you guys (especially Adam Beck and Zvi Shteingart) think about this idea in terms of the efficacy of this approach and the ease of implementation. Obviously, this wouldn't work well for people wanting to mine altcoins like peercoin, but a simple checkbox could fix that. Most miners are lazy and greedy, and just want to plug and play something that makes them money, which means they probably take care of the security of the blockchain unless their manufacturers do it for them. http://www.reddit.com/r/Bitcoin/comments/2873ox/making_pooled_mining_immune_to_51_attacks_selfish/Such logic can only be present in cgminer. I do not think this is practical or possible to put protection in the miner, stratum protocol assumes that the pool is playing by the rules. I do not think that it is in the interest of any pool to do 51% attack, since miners will notice this activity and leave the pool immediately. Also, big pools are a party most interested in the stability of BTC currency and have no commercial interest to destabilise it. Even without ghash.io, if you have 2 pools with 30% each no-one can promise that they will not do 51% attack together and share the profit. I think that stability of BTC in terms of 51% attack is assured by the game theory dynamics - I do not see how can anyone reaching capability to do 51% attack be interested in doing so.
|
|
|
|
zvisha
|
|
June 17, 2014, 12:45:30 PM |
|
You could see a lot of action (vapor bubbles) escaping through the mounting holes and out from under the edges of the board as the vapor builds up under the board it makes for an interesting show but obviously this is not the normal configuration. If it were a vertical setup you would have that rolling "boil" look I suppose. It just bet the SP10 board so well designed that in comparison to a smaller pcb of a higher density with 'older' hotter chips is comparing apples to dragonfruit with the SP10. They don't really need to be in Novec with those Hammer chips, but if you had a much smaller blade maybe that would be required. It be great to get some numbers on the SP10 especially can you overclock the snot out of the Hammer chips in the Novec even laying flat? Or do we all have to wait for Inside Bitcoin Hong Kong to learn more about the potential? I can't wait to get to Allied Control and check all these things out.
Curiosity!
Note that the FW pushes jobs to ASICs at the rate of ~600 jobs per second. If some ASICs run too fast (I think about 630Hz or more) they will finish job before the next job is ready, so they will run on idle. This gives you theoretical limitation of about 1.8T. You will see that the FW claims 1.9Th, but in pool you will see 1.8Th. If ASICs run faster then 630Hz, I will need to compile FW that pushes jobs at faster rate.
|
|
|
|
|